Science . . . generically
I was thinking. . .
Things are knowable insofar as they have actuality. All things have actuality, and are therefore intrinsically knowable in themselves. But to be knowable in itself is not necessarily to be known by this or that particular knower. Again, there is also knowledge of the thing, and one's knowledge of one's knowledge of the thing. [knowing vs. self-consciousness]
Turn to science - generically. We make statements about nature. This is knowledge. We make statements about our knowledge of nature - and this is epistemology. We make statements about our science (philosophically) and this is meta-science.
The confusion in modern science (specifically) is a level confusion between science, epistemology and meta-science. As shulamite pointed out, the only reason people read Dawkins et al. is because of the philosophy. Thus these books should not be called biology, but meta-biology. Examine quantum indeterminacy. Is that indeterminate because of the thing itself, or because of our knowledge of the thing (or perhaps both). To say that quanta have wave-like and particle-like properties is perhaps a difficulty in language - our epistemological understanding of electro-magnetism in modernity. Perhaps the thing itself in question has an actuality we have not fully described, and is therefore more knowable to us? Perhaps it lacks the actuality of other objects in nature, and is hence more potency (or less actualization of said potency) and thus our cursory knowledge is complete? Perhaps, epistemologically, the mathematical description of nature is not the complete language needed to describe the world? Perhaps, finally, we need to slow down and remember our A,B,Ds.